Genetic Parameters for Milk Yield and Milk Component Traits Estimated from Test Day and 305D Lactation Records of Philippine Dairy Buffaloes

Genetic parameters were estimated from test day and 305D Lactation records of multi-parity Philippine dairy buffaloes by random regression and repeatability models. There were 16,901 test day records from 2,303 lactations of 1st - 3rd parity cows from 1997 to 2012. Traits included milk, fat and protein yields, fat percentage and protein percentage. Varying orders of Legendre polynomials were combined with the Wilmink's function and were used in random regression. Heritabilities for lactation yields estimated from 305D records using a multivariate repeatability model were moderate; 0.25, 0.14, 0.19, 0.28 and 0.20 for milk yield, fat yield, protein yield, fat percentage and protein percentage traits, respectively. Heritabilities based on test day records using a multivariate repeatability model were lower but so were the standard errors. Heritabilities for individual test days of first parity cows estimated based on random regression models were higher than those obtained using repeatability models, ranging from 0.14 - 0.27 for milk yield; 0.11 - 0.14 for fat yield; 0.11 - 0.18, for protein yield; 0.12 - 0.33 for fat percentage and 0.05 - 0.29 for protein percentage traits. Genetic correlations between yield traits were highly positive whereas it was negative and moderate to high between milk yield and percentage traits. Thebest random regression model fitted a Wilmink's function for milk yield and a first order Legendre polynomial for milk component traits for random additive genetic and permanent environment effects. This model provides parameters suitable for selecting cows for milk production and persistency.